Initiative for Policy Dialogue Working Paper Series

October 2008

Flying Geese and Waddling Ducks: The Different Capabilities of East Asia

and Latin America to ‘Demand-Adapt’ and ‘Supply-Upgrade’ Jose Gabriel Palma

Industrial Policy

No part of this working paper may be reproduced or utilized in any form or by any means, electronic or mechanical, including photocopying, recording, or by information storage or

retrieval system, without permission from the Initiative for Policy Dialogue.

FLYING GEESE AND WADDLING DUCKS: THE DIFFERENT CAPABILITIES OF EAST ASIA AND LATIN AMERICA TO “DEMAND-ADAPT” AND “SUPPLY-UPGRADE” THEIR EXPORT PRODUCTIVE CAPACITY

José Gabriel Palma1 University of Cambridge

1

“The world of foreign trade is one of change. It makes a great difference to the trade of different countries, and to the impact of trade on them, whether they are capable of changing with the world. [...] Capacity to transform is capacity to react to change, originating at home or abroad, by adapting the structure of foreign trade to the new

situation in an economic fashion.”

-- C. Kindleberger

“[...] the obsession with competitiveness is not only wrong but dangerous, skewing domestic policies and threatening the international economic system.”

-- P. Krugman

Introduction

During the 1992 US presidential campaign President Bush's head economic adviser

was questioned about the decline in the technological content of US’s exports; he

replied that he saw no problem in it as there was no difference between exporting

micro-chips or potato-chips. This paper investigates whether this is really the case,

especially for developing countries (DCs). Are there major economic consequences

for the growth path of DCs (especially for its pace, nature and sustainability) from an

export orientation based on one or the other type of product? In particular, what are

the consequences in terms of the long-term productivity growth potential of the export

sector itself, for the capacity of exports to induce productivity growth in the rest of the

economy, for institution building, and for the welfare gains from specialization (e.g.,

issues related to the terms of trade)? Do DCs have a choice in this matter? And are

regional dynamics, particularly the role of the regional power, an important component

of the likelihood of DCs exporting one or the other type of product?

The general issue of “potato-chips versus micro-chips” will be analyzed

through the examination of the diverse export and growth performances of Latin

America (LA) and East Asia (EA) since 1960. As mentioned in the abstract, in this

chapter I argue that their experiences support at least four hypotheses. First, that

2

growth is a “product specific” phenomenon. Second, that the role played in it by

exports relates at least as much to what a country exports as to how much it exports –

i.e., that the capacity of exports to generate and sustain (trade-induced) GDP growth is

related not only to the volume but also to the composition of exports. Third, that the

capacity of an economy continuously to shift resources towards more “growth-

enhancing” export activities is related to the effectiveness of the state to implement

appropriate trade and industrial policies; and that this effectiveness is associated not

just with the ability of the state to create rents, but (much more importantly) to its

capacity to compel the corporate sector to invest them in productive capacity

diversification – i.e., continuously to shift resources towards products that would help

to supply-upgrade along the so-called “learning curve”, and demand-adapt a country’s

export productive capacity to an ever-changing international demand. And fourth, that

regional dynamics have played a significant role in growth, export diversification and

gains from specialization, especially due to the specific type of leadership that Japan

has exerted in East Asia (as opposed to that shown by the US in the Americas).

Economic growth: regional diversities

One of the main stylized facts of the world economy since the beginning of

‘globalization’ and economic reforms is the different economic performances of LA

and EA. Figure 8.1 highlights this remarkable phenomenon.

3

Figure 8.1

● GDP pc=gross domestic product per capita. Mex/China=GDP per capita of Mexico as a multiple of that of China. 3-year moving averages.

If between 1960 and 1980 Brazil practically doubled the gap between its GDP

pc and that of India, between 1980 and 2006 it was India’s turn to more than half this

gap. At the same time, while before 1980 Brazil was closing its income pc gap vis-à-

vis the countries of the OECD in a similar fashion to Korea, after 1980 only Korea was

able to continue to do this. And in terms of Mexico and China, the GDP pc of the

former as a multiple of that of latter collapsed between 1980 and 2006 from a factor of

28 to one of just 4.2

4

The primary objective of this chapter is to discuss some trade issues related to

these remarkable asymmetries and their regional dimensions.

East Asian versus Latin American development strategies

Differences between LA’s and EA’s trade and industrial policies have been well

documented. The primary benefit of these policies for EA (and now for India and

Vietnam) was that EA was able to increase simultaneously its shares of exports and of

manufactures in GDP. By contrast, LA, which throughout the first half of the 20th

century had been the region with the largest share of exports in GDP, between 1950

and the first oil crisis of the 1970s experienced a near-halving of this share. This

decline followed both a weakened demand for primary commodities in OECD markets

since the Korean War and trade and industrial policies characterized by strong anti-

export bias. As a result, and despite a strong growth performance in some countries

until 1980, particularly in Brazil and Mexico, the rapid increase in the share of

manufactures in GDP resulting from import-substituting industrialization (ISI) came at

the expense of exports.3

Although LA’s growth strategy during ISI had left the region vulnerable to

external shocks, especially due to continuous current account deficits and growing

foreign debt, nothing had prepared the region for the one unnecessarily created by Paul

Volker following his appointment to the Fed in 1979. In fact, the consequences for LA

of Volker’s “macho-monetarism” were of a magnitude not felt in the region since

1929.4 And as had happened in the 1930s, a massive and long-lasting external shock

(that found LA in an extremely vulnerable position) not only brought about the need

for a sudden and very painful internal and external macroeconomic adjustment, but

5

also laid the foundations for a radical and widespread change in the economic

paradigm of the region. In this case, it was characterized by an extreme move towards

trade and financial liberalization, wholesale privatization and market deregulation,

along the lines experimented with in Chile since 1973. Therefore, the key element to

understanding these reforms, particularly the purely ideological and remarkably rigid

and unimaginative way in which they were implemented throughout the region, is that

they were mostly carried out as a result of perceived economic weaknesses – i.e., an

attitude of ‘throwing in the towel’ vis-à-vis the previous growth strategy of state-led

ISI. In fact, the economic discourse ended up as a compass whose 'magnetic north'

was simply the mechanical reversal of as many aspects as possible of the previous

development strategy.5 The mere idea that alternatives could exist met with a mixture

of amusement and contempt.6 This helps to explain the remarkable differences with

which the reforms were implemented in LA, as distinct from EA.

As I have argued elsewhere (Palma, 2004a), it is not that EA did not implement

its economic reforms partly out of necessity (and also because of mounting external

political pressure from the US to do so); but its economic weaknesses were very

different in nature and intensity. Since the 1960s, EA integrated into the world

economy in a different way to LA. Instead of accepting their traditional, path-

dependent comparative advantages (i.e., based on traditional resource endowment,

subject to decreasing returns), East Asian countries acquired a more flexible and

growth-enhancing set of comparative advantages mainly by following a ‘flying-geese’

pattern of production and upgrading.7 This was achieved through increased export

penetration of OECD markets for manufactured goods, within a process of the

regionalization of production led by Japan (see especially Figure 8.6 below). Their

extraordinary success was based on several factors. On the external front, the fast rate

6

of expansion of international trade in these goods, and OECD (especially US) market

openness were key.8 Internally, it was due to several factors such as the ability to

build a structure of property rights and incentives, an institutional capability and a

political settlement that allowed them to produce globally competitive manufactures;

an ability to upgrade exports continuously through the ”flying geese” path;9 to their

ability to generate the high levels of investment and saving required for this upgrading;

and their achieving an effective coordination of this investment through different

forms of industrial and trade policies, which succeeded in simultaneously insulating

domestic markets and outwardly orienting tradable production.

However, in the late 1980s and early 1990s, problems emerged for many East

Asian economies. First, in part due to the increased standardization of inputs to the

electronics industry, some of their most important exports faced rapidly falling prices.

In response, the corporate sectors massively expanded their productive capacity,

attempting to turn falling prices to their advantage via increased market shares.10 An

obvious casualty was profitability; consequently, the composition of the finance for

investment had to move away from profits towards (domestic and foreign) debt.11 The

increasing need for access to finance was one of the key domestic pressures behind the

drive towards financial liberalization.

Another problem was that in the same period, China became a formidable

competitor in many markets crucial to the second-tier East Asian NICs, affecting

profitability and contributing to an increased need for external finance. At the same

time, the second-tier NICs reached a point where further upgrading of exports to

higher value-added products was becoming increasingly difficult. In particular, it

became more and more complicated to break away from a “sub-contracting” type of

industrialization. So (as India and Vietnam would do later), in a far more pragmatic

7

way than in LA, East Asian countries increasingly looked towards trade and financial

liberalization, and economic deregulation in general, not as a way of changing but of

strengthening their existing ambitious growth strategy.

The capacity to ‘demand-adapt’ a country’s export profile to a rapidly changing

OECD import-demand structure.

The rapidly evolving structure of demand in world markets is a well-known

characteristic of international trade, especially since the Second World War. This

phenomenon raises the issue of the need for a more flexible approach to the concept of

comparative advantages within an export-led growth strategy; i.e., from the point of

view of gains from specialization the key issue becomes how to develop the capability

continuously to adapt one’s export productive capacity to the ever-changing patterns of

international demand.

Since the introduction of the Prebisch-Singer hypothesis in the 1950s,12 and the

related research at ECLAC in the 1960s,13 DCs have fretted over the changing nature

of international demand, and in particular about the declining purchasing power of

unprocessed primary commodities exports resulting from their low income and price

elasticities of demand. Figure 8.2 shows the remarkable changes in the structure of

OECD demand for imports until 2000 (i.e., before the cyclical upturn for commodities

following 9/11) and its anti-unprocessed-primary-commodity bias (except for the

erratically changing fortunes of oil).

8

Figure 8.2

● Prim Com=primary commodities (SITC groups 0, 1, 2 and 4); Mf by Mat=Manufactures goods classified chiefly by materials (SITC group 6); fuels=Mineral fuels, lubricants and related materials (SITC group 3); Chemicals=Chemicals and related products (SITC group 5); Misc+oth=Miscellaneous manufactured articles and commodities and transactions not classified elsewhere (SITC groups 8 and 9); and Machinery=machinery and transport equipment (SITC group 7).

Note: The Trade-CAN program was originally developed by Ousmene Mandeng at ECLAC in the 1980s; it has been updated periodically by ECLAC and the World Bank. The data source is the UN Com-trade on-line database (which starts in 1963). Unless otherwise stated, this program is the source of all trade statistics quoted in this paper, and all export-groups quoted will correspond to the SITC classification, Version 2, at a 4-digit level.

● Source: Trade-CAN (2005).

Figure 8.2 shows the most important characteristic of the structure of OECD

imports between 1963 and 2000: the collapse of the share of non-oil primary

commodities (SITC groups 0, 1, 2 and 4). In fact, even including oil unprocessed

commodities lost about three quarters of their share in OECD imports during this 37-

year period, falling from 41% of the total in 1963 to just 10.6% in 2000 (left-hand

panel); and excluding oil from both the numerator and denominator, they fell from

46% to 11.5%, respectively (right-hand panel).14 However, primary products with a

higher level of manufacturing value added (e.g., SITC group 6) maintained their share

9

in OECD imports at a relatively stable level (about 15% of the total). By contrast,

imports of machinery and transport equipment (SITC group 7) more than doubled their

OECD import-share during this period (from 18.4% to 41%).

Quite apart from supply-side issues discussed below (such as the different long-

term productivity growth potentials that exports of manufactures and unprocessed

commodities may have), to insert oneself into the international division of labour

through exporting products characterized by slow-growing (or even declining) markets

does not augur well for long-term welfare gains from international specialization. In

fact, during the second half of the 20th century, the terms of trade of non-oil exporting

LA fell (cyclically) by half from its Korean War peak, reaching before the 9/11 crisis a

level even below that of the post-1929 crash. What Latin American policy-makers and

their economic advisors seem to forget these days is that, from the point of view of

demand, exports are just an indirect way of producing imports; and by exporting low-

income and low-price elasticity of demand products, declining terms of trade makes

(the indirect production of) those imports ever more expensive. In fact, in a Ricardian

sense, the region’s increased export efficiency may sometimes even act as an ‘own-

goal’ because although increased competitiveness may help GDP-growth, if it is done

too much at the expense of the terms of trade, it may seriously harm the purchasing

power of that GDP growth – i.e., in extreme cases output growth may be outweighed

by deteriorating terms of trade.15

As if the problem of the (cyclical) long-term declining level of LA’s non-fuel

terms of trade was not serious enough, the instability of these terms of trade adds a

further uncertainty to the region’s economic life. In fact, the average (absolute)

percentage variation of LA’s non-oil terms of trade reached 6.7% per annum during

the 84-year period between 1920 and 2006. To give an indication of the magnitude of

10

this instability problem (and assuming an average share of exports in national income

of one third, with everything else constant), a country facing this amount of instability

in its export markets will have its national income changing randomly by an average of

2.2% per annum for this reason alone!

However, demand dynamics in international markets is a more complex issue

than just a generic one between unprocessed commodities or manufactures. There is

also a significant diversity within manufactures in this respect.

FIGURE 8.3

● Industries are grouped on the basis of their R&D intensity. Cars=road vehicles, engines and their non-electrical parts; medium R&D*=excludes vehicles and engines.

● Source: Trade-CAN (2005), using the OECD classification for product content of R&D.

11

Figure 8.3 illustrates that the rapid growth of OECD demand for manufactures

is concentrated in just two types of product: one requiring intense R&D and one of

moderately intense R&D (road vehicles, including their engines and non-electrical

parts). These two groups had a sector-share increase between 1963 and 2000 of 317%

and 160%, respectively. Sector shares of other manufactured products with a medium

content of R&D increased by only 14% during this 37-year period, while that of

manufactured products with a low content of R&D actually declined by 23%. In turn,

the market-share of resource-based products collapsed by 71%! Figure 8.3 also

reveals the remarkable correlation between demand dynamism and technology content

of OECD imports. Therefore, the degree of dynamism of OECD demand for imports

of a product could also be understood as a proxy for the technology content of that

product (see Figure 8.10 below).

Given the fact that OECD demand for imports is a rapidly moving target, one

of the key differences between LA and EA during this period was that while EA was

able to develop its skills at keeping an eye on this moving target (and kept shifting

resources accordingly), LA continued aiming as though at a fixed target (increased

competitiveness in non-demand-dynamic products; see Figures 8.4, 8.9 and 8.10

below). In fact, one could argue that the real comparative advantage acquired by EA

during this period was not on a set of products, but on learning how to develop a

flexible export productive capacity.16

Writing at the beginning of the period studied in this paper, Kindleberger

(1962, p. 10; see introductory quotation) predicted this rapidly changing structure of

world trade, and prescribed developing “the capacity of changing with the world” as a

necessary skill for maximizing gains from trade. Figure 8.4 presents a statistical

indicator for a country’s capacity to react to changes in world demand via a “demand-

12

adaptability index”.17 This index is constructed using the ratio of a country’s or

region’s market share in “demand-dynamic” sectors (adjusted by the weight of

demand-dynamic sectors in all sectors of OECD imports), to that of its market share in

“demand-sluggish” sectors (adjusted by these sectors’ share in OECD imports). A

value of 1 for this index at the end year of a period indicates that during the preceding

years a country was able to “track” changes in demand in OECD markets – i.e., it was

able to ‘demand-adapt’ its export structure to reflect the changing structure of OECD

imports.18

Figure 8.4

● Excludes oil. M=imports; Jap=Japan; N-1=first-tier NICs (Korea, Singapore and Taiwan; Hong Kong has been excluded due to its transformation into a financial centre); N-2=second-tier NICs (Malaysia and Thailand); LA*=LA (excluding Mexico and Brazil); Bra=Brazil; SA=South Asia (India and Pakistan); S-SA*=Sub-Saharan Africa (excluding South Africa); ZA=South Africa.

Note: The index for Mexico is not included in the graph due to the distorting effect of its ‘maquila’ exports, which in trade statistics are measured in terms of gross value of production (i.e., including imported inputs); ‘maquila’ exports accounted for about half its manufacturing exports, but almost 80% of the value of output was made from imported inputs (see Palma, 2005a). The share of imported inputs is also very high in the rest of manufacturing exports (see Capdevielle, 2005; and Moreno-Brid, 2005).

13

Figure 8.4 shows LA’s tendency to under-invest in productive capacity

diversification in its export sector; this could hardly be passively blamed merely on the

fact that the region happens to be rich in natural resources. Abundance of natural

resources could hardly be considered a sufficient condition for lack of investment

efforts in product-diversification and upgrading. That is, a rich resource endowment

could not be blamed in itself (as the resource-curse hypothesis seeks to do) for a

country ending up being a waddling-duck rather than a dynamic goose.19

One obvious way for primary commodity rich countries to improve their

‘demand-adaptability’ is to increase the degree of processing of these commodities, as

the Scandinavian and some East Asian countries (especially Malaysia) have shown.

Another is the upgrading of their domestic productive capacity in terms of the inputs

needed for their commodity production.20 Both paths (up- and down-stream) would

lead to export products that are more “demand dynamic” in world markets. Yet

another path is the targeting of high-value low-volume commodity niches, taking

advantage of the hyper-segmentation of markets that characterizes the current process

of globalization.21 The process of export upgrading (both up- and downstream) is the

subject of the next section of this paper.

The capacity to “supply-upgrade” a country’s export profile

Upgrading resource-based exports

Increasing the degree of manufacturing value of resource-based exports allows

14

resource-rich countries to attain an export structure characterized not just by stronger

demand in international markets, but also by products with higher long-term

productivity growth potentials. Figure 8.5 shows the contrast between Finland and

Malaysia on the one hand, and Brazil and Chile on the other, in timber-based export

products.

Figure 8.5: Finland, Malaysia, Chile and Brazil: vertical integration in the processed-timber industry

● w-chip=wood-chips; pulp=pulp and waste paper; w-maf=wood manufactures; furnit=furniture; w-mach=paper mill and pulp mill machinery. Finland also managed a substantial share in OECD imports of paper and paperboards (not included in the graph), reaching in 2000 a market share in OECD markets of 14% (in that year, the respective market shares for Brazil and Chile were just 0.6% and 0.1%).

The differences in long-term productivity growth potentials characteristic of

15

processed and unprocessed timber are well known. There are also substantial

differences in their demand-dynamics in OECD import markets: while the sector-share

in total OECD imports of wood-chips fell by 54% and that of pulp by 71% between

1963 and 2000, those of wood manufactures increased by 74% and of furniture by

400% during this period.

As Figure 8.5 shows, Finland’s high market shares in OECD imports of the

more technologically advanced timber-based products were acquired gradually as they

moved along the value-added learning curve of the industry. A large part of Finland's

timber-based exports in the 1960s consisted simply (as in Brazil and Chile) of wood-

chips and pulp. However, by the 1970s Finland had all but stopped exporting wood-

chips, had substantially cut its exports of pulp, and was using these raw materials as

inputs for increased production and export of paper and paperboards, wood

manufactures, and furniture. Also, by then Finland’s market shares in OECD markets

for machinery for the paper mill and pulp mill industries had exceeded 8% – a

substantial achievement for such a small country.22

Figure 8.5 also reveals large disparities in export upgrading in the processed-

timber industry between LA and EA. In the 1960s, neither had a processed-timber

export industry. In the 1970s, while Chile and Brazil continued to export basic timber

products, Malaysia had already added some wood manufactures, aided by the

successful implementation of an industrial policy restricting exports of unprocessed

timber. By 2000, Brazil’s and Chile's exports still consisted entirely of products at the

lower end of processing, while Malaysia's exports of processed-timber products

consisted almost entirely of wood manufactures and furniture.23

The upgrading in the timber-based industry is, of course, a complex

phenomenon, full of investment indivisibilities, high-entry costs, huge economies of

16

scale, intricate financial engineering, specific skills, technological complexities and

environmental issues. Even the relatively simple operation of upgrading from wood-

chip to ‘MDF’ requires considerable investments and skilful institutional

arrangements.24 Basically, one could easily end up in the business of transforming

high quality timber into worthless furniture. As a result, many advise DCs not to

tackle these market failures in the export-upgrading business (those that tend to keep

them trapped in their traditional comparative advantages), and to try to make the best

of their sub-optimal growth path. It would be wiser to be risk averse in this matter of

‘acquiring’ more growth-enhancing comparative advantages. For example, some may

remember the World Bank reports and academic publications of the late 1950s and

early 1960s that were intrigued by Japanese corporations transforming world-class

steel and other high-quality inputs into sub-standard cars, which could be sold only to

captive domestic customers or exported at highly-subsidized prices.25 Korean

corporations followed suit in the 1970s, and the ‘usual suspects’ again wondered about

what could be the rationale for building low or negative value-added production lines

that used profits from captive markets to subsidize unprofitable exports? Who would

voluntarily choose to buy a 1970s’ Korean car, unless it was a home-customer with no

choice, or an international one only able to afford a massively subsidized product? But

in fact, it did not take all that long for the Japanese and Korean car industry to turn the

tables completely on their competitors.26

Asymmetries in the dynamics of international demand for iron, steel and metal-

working machinery are similar to those of the timber industry: while the sector shares

in total OECD imports of iron ores and concentrates fell in value by 82% between

1977 and 2000, those of metal-working machinery significantly improved their relative

position in OECD imports. However, in 2000 Chile exported only iron ores and

17

Brazil’s market share in OECD imports of iron ores was 15 times larger than its share

in processed iron and steel. As mentioned above, this can hardly be justified simply by

LA’s natural resource abundance or US protectionist policies. Korea, by contrast,

soon transformed the initial exports of iron ores into steel; then developed further into

exporting metal-working machinery. Taiwan, in turn, had by the 1980s also already

developed a significant export industry of metal-working machinery.

Although it is well known that as many countries which have tried this type of

product-upgrading and demand adaptation have failed as have succeeded, the fact is

that EA has developed the habit of succeeding while post-reform LA has stopped

trying. Not surprisingly, when FDI decides to get involved in this type of value-chain

it tends to invest in resource-extraction in LA and resource-processing in EA.27

In fact, post-reform LA has not only invested very little in productive capacity

diversification in order to move up the ‘technology ladder’ in terms of the processing

of exports, but in several cases the movement has actually been in the opposite

direction. Post-1973 Chile, for example, not only abandoned its previous ‘pro-

industrialization’ agenda, but even ended up severely reducing the share of domestic

value added in its copper exports, with the proportion of refined copper in total exports

being drastically reduced in favor of the far more primitive copper ‘concentrates’ (see

below).

Supply-upgrading in resource-poor countries: the flying-geese phenomenon

The Japanese economist Akamatsu coined the phrase ‘flying geese’ to characterize the

East Asian supply-upgrading industrialization model (Akamatsu 1962). The analysis

that follows will characterize two very distinct components of this flying-geese

18

phenomenon, not distinguished properly by Akamatsu and often confused in the

literature: what I have called here the “sequential movements” and the “parallel

movements” along the learning curve. Although each process is characterized by a

very different dynamic, both have a crucial common element in that nearly all products

involved are demand-dynamic from the point of view of world trade. What is

characteristic of the first (and better known) component of this ‘flying-geese’ process

of industrialization, that which ‘moves sequentially along the learning curve’, is the

involvement of products that Japan can no longer competitively produce and export

and, therefore, allows its productive capacity to be transferred to the geese that are

following them. For Japan, these products are either too labour-intensive to be

produced competitively at Japanese wages, or they have already exhausted their

productivity growth potentials (Figure 8.6, left panel).

Figure 8.6

● Panel A=percentage of exports in the respective countries in Japan’s 10 export sectors with most rapidly declining market shares between 1963 and 2000 (SITC groups 843, 851, 037, 034, 894, 899, 666, 842, 761 and 762; for a detailed analysis of these products, see Palma, 1998). Panel B=the

19

same for Japan’s 20 major export sectors in 1995 (SITC groups 781, 752, 764, 784, 776, 759, 778, 713, 763, 894, 782, 785, 751, 772, 749, 898, 874, 882, 728 and 736). MA+Th and M&T=Malaysia and Thailand; and Ko=Korea.

In the “sequential-movement” component, export productive capacity of products that

tend not to be very high up the ‘learning curve’ is successively transferred from Japan

to the first-tier NICs, and then from these countries to the second-tier NICs, then to

China, Vietnam, and so on. The crucial issue here is that when productive capacity is

transferred to the geese that are following them, these substitute the exports of the

more advanced goose in world markets, creating a sequence of ‘inverted U-paths’ – a

phenomenon widely acknowledged in the traditional flying-geese literature.

The second (and usually ignored) component of the flying-geese phenomenon,

that of “moving in parallel” along the learning curve (Figure 8.6, right panel), reflects

a different regional dynamic. This relates to products in which Japan does not

relinquish its productive capacity easily and often vigorously fights back against the

challenge of other Asian countries. Japan’s reluctance to ‘concede’ is due either to the

fact that these products still have substantial productivity growth potential left in them

and/or because they are among the products with the highest demand-dynamics in

world trade.28

As Figure 8.6 indicates, Japan’s share of exports in these 20 sectors eventually

began to decline under the competitive pressure of the other East Asian geese; even in

these extremely demand-dynamic markets, there is limited space for market entry. So,

here we find not ‘inverted U-paths’, but parallel trajectories; the crucial issue here is

that as this process involves products that are much higher up in the ‘learning curve’.

These are products in which Japan is reluctant to give up, and East Asian countries

20

(with lags in time associated to their relative level of development) attempt to compete

with Japan (rather than substitute its exports, as in the ‘inverted-U’ case).

One of the first casualties of EA’s moving ‘in parallel’ along the learning curve

was the US; this was due to the fact that this pattern of “imitation-with-a-lag” leads to

an incremental East Asian market-share (see right-hand panel of Figure 8.7). This is

not the case with the other component of flying-geese industrialization, the ‘sequential’

movement in less attractive export products, as EA’s aggregate market share hardly

increased as productive capacity was systematically moved from one country to the

other.

Figure 8.7

● LA-M=Latin America, excluding Mexico; East Asia=combined market shares of Japan, the NICs-1, NICs-2 and China in Japan’s 20 major export sectors in 1995. (For a list of these export sectors see

21

Figure 6.)

The contrast between EA and the Americas in both components of the East

Asian ‘flying-geese’ pattern of industrialization is remarkable. On the one hand, in

terms of Japan’s declining sectors, the US and non-Mexico-LA maintain remarkably

low and stable market shares for most of the period (left-hand panel in Figure 8.7).

Yet, on the other, during this period the US lost half its market share in OECD imports

in the second (more growth-enhancing) group of products (right-hand panel in Figure

8.7). In turn, LA (excluding ‘maquila’ Mexico) is on a completely different

productive-capacity planet as far as these products are concerned.

However, to test whether or not there is a ‘flying-geese’ phenomenon in the

Americas one should compare LA’s exports with those of the US, not just with those

of Japan. Figure 8.8 does that vis-à-vis the US’s 20 major and most ‘dynamic’ export

sectors.29

Figure 8.8

22

● X=exports. US’s 20 major export sectors in 1995 were: SITC groups 792, 752, 784, 759, 781, 776, 874, 714, 764, 713, 931, 541, 778, 898, 772, 598, 749, 222, 322 and 892. US’s 20 most ‘dynamic’ export sectors (i.e., those which had the fastest increase in market shares) 1963-2000: SITC groups 246, 785, 034, 289, 112, 035, 061, 323, 014, 634, 036, 659, 248, 001, 871, 072, 223, 247, 287, and 291 (to avoid spurious growth rates in the calculation of US’s 20 most ‘dynamic’ export-sectors, a ‘filter’ was used to include only those export sectors that had an export-share of at least 0.3% of the total in 1963).

Within LA, Brazil was the only country that had a slow but steady increase of

the share of its exports in the US’s ‘top-20’ sectors (right-hand panel in Figure 8.8),

reaching 22% of its exports in 2000; however, more than one fifth of this increased

share is made up by a single product, and a primary commodity at that: soybeans! The

rest of ‘non-maquila’ LA had no significant amount of export-shares in these products

(though Argentina was also rapidly increasing its soybean exports towards the end of

this period).

One could hardly characterize this as a regional flying-geese pattern of

industrialization and upgrading, one in which exports follow either a sequential or a

parallel path between the US and LA. Of the US’s 20 most dynamic export sectors

(left-hand panel in Figure 8.8), 16 are primary commodities and fuels. Not

surprisingly, the combined sector-share of these 20 sectors in all OECD imports fell

steadily from 10.4% in 1963 to just 4.4% in 2000. Contrast this with Japan's 20 most

dynamic export sectors, which doubled their sector-share over the same period.

Moreover, as the right panel of Figure 8.8 shows, in the US’s 20 most dynamic export

sectors the “goose leader”, the US, flies in reverse: i.e., trying to penetrate markets that

are being relinquished by supposed geese-following Latin American countries! That

is, it is in the 20 sectors in which the US had the highest increase in OECD market

shares that one of the most bizarre aspects of the US’s regional influence emerges. The

US was increasing its export-shares in (rather unattractive, mostly primary) products in

23

which LA was already reducing its export-share. Rather than an East Asian flying-

geese pattern of succession, the (more magical realist) Americas present a scenario in

which the leader follows the followers…30

Figure 8.9, which compares the changes in the overall levels of competitiveness

and the capacity to move into ‘high-tech’ products in the regions studied here,

illustrates several related issues.31

Figure 8.9

● ‘High-tech’ products=products with high content of R&D. X=exports and M=imports. Chi=China; Latin America*=Latin America, excluding Brazil and ‘maquila’ Mexico and Central America; Tw=Taiwan.

● Vertical axis: percentage of exports with high content of R&D in 1963 (first observation) and in 2000 (second observation). Horizontal axis: percentage of all exports (with or without high content of R&D) in which the respective country or region gained market shares in OECD imports between 1963 and 1971 (first observation), and between 1990 and 2000 (second observation).

24

First, Figure 8.9 shows that in 1963 neither LA nor the first- and second-tier

NICs and China exported ‘high-tech’ products. However, secondly, it also indicates

that East and South East Asian countries were already competitive in the (non-high-

tech) products that they exported at the time (products that were rather similar in

nature to those exported by LA). Third, it shows that among Latin American countries

during the 1960s, only Brazil and Mexico (not included in the graph) were relatively

competitive in their exports and were gaining market shares in a significant proportion

of them (58% of exports in both countries); i.e., LA’s ISI-anti-export-bias affected

more intensively the export-competitiveness of the small and medium-sized countries

(including Argentina). Fourth, it shows that during the 1990s (after trade

liberalization) LA was able massively to increase its competitiveness in OECD

markets – in several cases, from less than 20% to about 80% of their exports. Fifth,

even though during the 1990s LA reached East Asian levels of export-competitiveness,

it did so only in its traditional-type exports.32 If international demand had not

discriminated in such an extreme form against commodities, and if these products had

had the same long-term productivity growth potential and growth-enhancing ‘pulling’

effect as those higher up in the ‘value-chain’, this remarkable increase in

competitiveness would have been an unqualified regional success. Unfortunately, this

was clearly not the case. Finally, Figure 8.9 also illustrates East Asian countries’

remarkable capacity to ‘reinvent’ themselves, by diversifying their export productive

capacity towards ‘high-tech’ products; further, they did so while retaining most of their

former extremely high levels of export-competitiveness.

Overall export-trajectories between the 1960s and 1990s

Finally, one could compare export profiles between the 1960s and the 1990s not just

25

by looking at the capacity to move into ‘high-tech’ products (Figure 8.9), but also into

‘demand-dynamic’ products in general, whether ‘high-tech’ or not (Figure 8.10).

Figure 8.10

● Excludes oil. X=exports; M=imports. Br=Brazil; Ch=China; EU=European Union; LA*=LA (excluding Brazil; maquila Mexico and Central America); Mal=Malaysia; N-1=Korea, Singapore and Taiwan; SSA=Sub-Saharan Africa (excluding South Africa); Thai=Thailand; Viet=Vietnam.

● First observation: export profile of country or region 1963-1971. Second observation: that between 1990 and 2000. Vertical axis: percentage of exports of a country or region of products that became ‘demand-dynamic’ in OECD imports (i.e., products that increased their share in OECD imports during respective periods due, for example, to their higher income elasticity). Horizontal axis: percentage of exports in which the respective country or region gained market shares in OECD imports during the relevant period. Therefore, if an observation is in Quadrant 1 this indicates an ‘uncompetitive’ country (i.e., less than half its exports have gained market shares) exporting ‘non-demand-dynamic’ products (i.e., less than half its exports are ‘demand-dynamic’ products); if it is in quadrant 2, it shows a ‘competitive’ country exporting ‘non-demand-dynamic’ products; if in quadrant 3, a ‘competitive’ country exporting ‘demand-dynamic’ products; and in quadrant 4, an ‘uncompetitive’ country exporting ‘demand-dynamic’ products.

Notes: i) Data for Taiwan correspond to those reported in the second edition of the Trade-CAN software; this edition includes data only until 1995. The third edition (used for all other countries) does not provide data for Taiwan; ii) The first observation for Vietnam corresponds to the period 1973-1984 (i.e., from the date when US combat troops left Vietnam until the beginning of economic reform; Trần Văn Thọ, et al. 2000); the second observation is for the 1990s; and iii) See Appendix 2 for a more formal definition of the four quadrants.

Figure 8.10 illustrates that the remarkable increase in export-competitiveness in

26

LA between the 1960s and the 1990s (i.e., the rapid movement from quadrants 1 to 2),

was not accompanied by an improvement in the ‘quality’ of its exports. Figure 8.9 has

shown that LA’s improved export-competitiveness did not include many high-tech

products; Figure 8.10 indicates that it did not include many other demand-dynamic

products either. In EA, in contrast, it indicates an intriguing aspect of the flying-geese

pattern of industrialization and exports: the different trajectories of young-

industrializers (second-tier NICs and China), to those of middle-aged ones (first-tier

NICs) and of more ‘mature’ ones (Japan). Basically, the rapid movement into

demand-dynamic products of the second-tier NICs and China – their sharp movement

from quadrants 2 to 3 – eats away some degree of export-competitiveness in the first-

tier NICs. This process is more acute with Japan (and the EU) because the aggregate

competitive pressure of all NICs and China (in addition to other domestic economic

problems) has driven these countries from quadrants 3 to 4. With the exception of the

US (primarily during the Clinton years, and in part due to the composition of its

exports discussed above), the overall pattern that emerges in Figure 8.10 is one of a

clear anti-clockwise trajectory.

For LA and other countries moving into quadrant 2, the crucial trade- and

industrial-policy issue next is whether there are endogenous market forces that would

lead them in an upwards movement from quadrants 2 to 3. Or whether there are

market failures that would lead them to get trapped in being ever more competitive in

products that tend to be ever more marginalized (in value terms) from world markets

(except for temporary cyclical periods such as those that benefited many commodities,

such as oil, after the 9/11 crisis). In fact, especially in commodity markets, excessive

competitive struggle for market shares by DCs often leads to a self-defeating fallacy of

composition problems.33

27

So far, there is little (if any) evidence of “upward endogenous forces” at work

that would help bring a country from 2 to 3. Countries in quadrant 2 seem to need an

exogenous push such as the trade and industrial policies that characterized the

movement from 2 to 3 in most East Asian countries. However, for these policies to be

implemented effectively, what is needed is the type of institutional arrangements that

allowed them to be effective in EA. These include a strong state, capable not only of

devising trade and industrial policies to generate rents that would create incentives to

the transfer of resources towards demand-dynamic export products, but also capable of

making sure that these rents are used effectively. That is, a state not only able of

generating rents, but also one capable of imposing conditionalities in order to

‘discipline’ the capitalist élite to use them productively. Furthermore, for these

policies to succeed it is also necessary to have a state capable of withstanding

clientelist pressures from “intermediate classes”.34

If these policies and the institutional arrangements necessary for their success

are not implemented, the potential GDP-growth-enhancing effect of further increases

in export-competitiveness in LA would continue to be restricted by the generally low

productivity growth potential of unprocessed commodity exports (sometimes after an

initial one-off boost), and by their modest positive externalities and spill-over effects

and low capacity to induce productivity growth in the rest of the economy in general.

Furthermore, lack of an upward movement from 2 to 3 could also seriously affect the

welfare gains from trade specialization in terms of the purchasing power of exports.35

Existing evidence for LA indicates that the (not so) invisible hand of globalized

markets are only creating incentives leading towards further penetration into quadrant

2. There is little evidence that market incentives, the domestic structure of property

rights and institutional arrangements, or domestic political settlements are helping the

28

region to move upwards in the direction of quadrant 3. One example is Chile, the most

successful country in post-reform LA (at least for the period between the 1982 debt

crisis and the 1997 East Asian financial crisis), whose export trajectory is a horizontal

(in fact, slightly downward) movement from quadrant 1 to 2, increasing the share of

competitive exports from 12% to 79%, while actually decreasing the share of exports

of demand-dynamic products from 12% to 6%. The performance of its copper export

industry is a good example of this peculiar combination of increased competitiveness

with no ‘upward push’; in fact, while rapidly gaining market share, Chile was actually

reducing the average share of manufacturing value added in its copper exports. As a

result, the proportion of refined copper (i.e., with more than 99% metal content) has

fallen from 97% of total copper exports in 1973 to just about half, in favor of the far

less processed form of copper ‘concentrates’ (with less than one-third metal content).36

There is ample evidence that the sharp slowdown in Chile’s growth since the end of

the 1990s is partly due to this under-investment in upward productive diversification.37

Finally, the analysis of the past 40 years of international trade clearly shows

that LA’s long-delayed transition towards quadrant 3 can expect little (if any) pulling

help from the US (as East Asian countries had from their ‘leading-goose’).38

A simple test for the different capabilities of EA and LA to use exports effectively

as an engine of growth

LA’s economic performance since the beginning of trade and financial liberalization,

and the switching back of the engine of growth to the export sector, has been

characterized by dynamic exports and sluggish growth in the rest of the economy.

Figure 8.11 shows the average growth of GDP and exports since 1980, as well as the

29

conditional expectation for this relationship.

Figure 8.11

● EA*=NICs-1 and NICs-2; LA*=Latin America (excluding Brazil, Mexico and Venezuela); dLA=interactive dummy for Latin American countries; mena=North Africa and the Middle East; oecd=average of 15 high-income OECD countries; SA*=South Asia (excluding India); S-S A*=Sub-Saharan Africa (excluding South Africa). Br=Brazil; Mx=Mexico; and ZA=South Africa.

● Source: Constructed by the author using data from WDI (2007).

Note: The regression line is obtained from a cross-country regression in which GDP growth is the dependent variable and export growth squared is the explanatory one; the sample consists of all countries for which the WDI reports data for both variables during this period (94). It is important to emphasize here that the aim of this regression is simply to represent a cross-sectional description of cross-country GDP-growth differences, categorized by export-growth performance. Therefore, this regression should not be interpreted as ‘predictive’, because as is well-known there are a number of difficulties with a regression of this type estimated from a single cross-section, especially regarding the homogeneity restrictions that are required to hold. This regression passes all the relevant diagnostic tests. The R2 is 54% and the ‘t’ statistic of the explanatory variable is 10.1. The ‘t’ statistics are based on ‘White’s heteroscedasticity adjusted standard errors’. For a discussion of the econometrics of cross-section regressions, see Pesaran, et al. (2000).

In most of LA, as in many other DCs, economic reform has certainly succeeded in

shifting the ‘engine of growth’ towards the export sector. However, for any engine to

be effective, the power it generates must be properly harnessed. Asian countries seem

30

to have mastered this process rather well, while those in LA (especially Brazil and

Mexico, formerly the most dynamic economies of the region) and in Sub-Saharan

Africa (including South Africa) remain far from doing so. Figure 8.12 shows the

degree to which the performance of the export engine seems to be region-specific, and

the remarkable underperformance of LA.

Figure 8.12

● EA=East and South-East Asia; LA=Latin America; dLA=interactive dummy for Latin American countries; SA=South Asia; S-S A=Sub-Saharan Africa; and S-S A*=Sub-Saharan Africa (excluding South Africa). Ar=Argentina; B=Bangladesh; Br=Brazil; Ch=Chile; Co=Colombia; Ec=Ecuador; H=Hong-Kong; In=India; Ko=Korea; M=Malaysia; Mx=Mexico; Pa=Paraguay; Pe=Peru; Pk=Pakistan; Th=Thailand; Tw=Taiwan; Ur=Uruguay; and VN=Vietnam; and ZA=South Africa. Vietnam’s export growth is 14.7%. With the LA dummy the R2 of the regression increases to 58% and the ‘t’ statistic of the explanatory variable to 10.8; the ‘t’ of the dummy is 3. The ‘t’ statistics are based on ‘White’s heteroscedasticity adjusted standard errors’.

Note: The WDI database does not provide information for Taiwan. Data for this country have been obtained from The Republic of China Yearbook of Statistics.

As is clear from the graph, the Latin American export-engine tends to perform badly

31

compared to those of EA and South Asia in terms of its capacity to be associated with

GDP growth in the rest of the economy. In fact, it performs poorly in relative terms

even when compared with that of other commodity-exporting countries. Basically,

during this period Sub-Saharan countries achieved disappointing growth rates on both

exports and GDP, while Asian countries managed dynamic growth on both fronts;

whereas Latin American countries (and especially Brazil and Mexico), while able to

engineer a rapid expansion of exports, were unable to pull the rest of the economy

along with it. So, in statistical terms, as GDP growth in Latin American countries is

located below their conditional expectations consistently enough, the interactive

dummy for the region ends up being negative and highly significant.

The underperforming of the Latin American export-engine after economic

reform is a relatively stable phenomenon throughout the post-1980 period. In Brazil,

for example, while between 1950 and 1980 the growth of the previous engine

(manufacturing) and overall GDP were very similar (8.8% and 7.3% per year,

respectively), since Lula was first elected, the asymmetry between the growth of the

new engine (exports) and GDP during this period (2002-2007) could hardly be greater

(13.3% and 3.5% respectively; see http://www.eclac.org).

Mexico is the country with the largest gap between actual GDP-growth and the

conditional expectation of this growth in the whole sample, making it an extreme

example of export-led failure. During these two and a half decades, a remarkably

dynamic growth of exports (9.2% per year) was associated with an extremely poor

growth performance (2.6%).39 Further, as population grew at over 1.6%, per capita

GDP growth stood at just under 1% per year.40 Figure 8.13 shows more clearly this

phenomenon of booming exports with little GDP-growth “tracking”.

32

Figure 8.13

● Log scales. Exports do not include oil; 1980 prices. For the convenience of the graph, before transforming the data into logs the original data were first made into a 3-year moving average and then into index-numbers with 1981 as a base year (equal to 20.09; the natural logarithm of this number is 3).

● Sources: 1950-2004 from Palma (2005a); 2005-2010: estimates and forecasts of the Economist Intelligence Unit.

The Mexican economy’s remarkable inability to harness the power generated by its

dynamic export expansion seems to be its most important economic failure since the

end of the ISI-period and the beginning of neo-liberal economic reforms. As Figure

8.13 shows, the long-lasting close relationship between exports and GDP growth, built

during ISI, disappeared after the 1982 debt crisis, when the sharp acceleration in the

rate of growth of exports became associated with a sharp decline in rate of growth of

GDP – not exactly the Promised Land of the Washington Consensus. In fact, the ratio

of export growth to that of GDP more than trebled.

33

As the regression line above has indicated, in LA exports are faltering as an

engine of growth not because of their lack of own growth, market penetration or

competitiveness, but because of their poor ‘GDP-tracking’ properties. This is likely to

be associated, at least in part, with the lack of ‘quality’ of its export composition (i.e.,

in terms of the idiom of Figure 8.10 above, with the region’s inability to move from

quadrants 2 to 3; and in the case of Mexico, with the lack of ‘deepening’ of its maquila

exports). Orthodox and unimaginative monetary, fiscal and exchange-rate policies

have not helped either. However, as Figure 8.12 indicated, this is also a region-

specific phenomenon, likely to be the result of the rigid way in which economic

reforms were implemented in the region. For example, if one adds to the above

regression a variable for the ‘quality’ of exports (as in Figure 8.10), other variables to

account for factor accumulation (population growth and investment) and others to

account for ‘production frontier shifters’ (such as manufacturing), LA again ends up as

the region with the largest negative gap between actual GDP growth and the

conditional expectation of that growth.

Ironically, while during the ISI period the main obstacle in LA to the

sustainability of growth was an almost obsessive concentration of policy incentives

and resources in (non-export) manufacturing, leading to a neglect of exports (hindering

growth via a balance of payments constraint), in the post-1980 period the region did

exactly the same but the other way around. Though LA certainly succeeded in

becoming competitive in world markets for its traditional export products, this effort

seems to have had little overall growth effects. Lack of growth-enhancing policies,

institutions capable of implementing them, adequate structures of property rights and

incentives and macro-stability also matter in this. Product diversification, high levels

of investment and adequate institutions seem to be the conditions that have helped

34

Asian countries successfully take advantage of new ‘growth spaces’ in a globalized

world.

In sum, export-led growth when based on unprocessed primary commodities or

‘thin’ maquila exports has proved to be a poor engine of growth. The main lesson

from post-reform LA is that if middle-income DCs want to insist on this export

orientation (or if there is little they can do about it in the short-term), policy makers

and the capitalist élite of these countries should think about this model only as an

“growth-enabling” export-strategy, not as an “export-led” growth strategy. That is, a

fast rate of growth of (unprocessed) commodity exports can at best be expected to

provide foreign exchange to enable a more dynamic rate of growth in the rest of the

economy; i.e., to increase the rate of growth of GDP that is sustainable from the point

of view of the balance of payment. However, for this growth to actually take place

there is still the need for a proper ‘engine’ to be activated somewhere else in the

economy. That is, parallel to commodity exports there is the need for other sectors or

activities to play the role of “production frontier shifters”, able to set in motion

processes of cumulative causation – characterized by their positive feedback loops into

the system, capable of generating a momentum of change which is self-perpetuating

(as in the Veblen/Myrdal or Young/Kaldor manner). Not much evidence from LA to

indicate that unprocessed primary commodities or ‘maquila’ exports can play such

role!

Conclusions

Ultimately, the effectiveness of exports as an engine of growth depends on the capacity

of exports to improve both productivity growth and labour absorption not only in the

35

export sector but in the overall economy. On the purchasing power of income side

there is the additional issue of the terms of trade, as these are the vital link between

real GDP growth and real command-GDP growth.

In post-reform LA (with the sole exception of Chile – and only during a

specific period that covers about half its post-reform years), the main stylized fact

regarding GDP growth is that even in those countries successful in rapidly accelerating

the rate of growth of exports, especially Mexico and Brazil, the rate of growth of

output has been particularly disappointing. Of course, for those countries that have not

even been able to improve export-competitiveness, the post-reform economic situation

has usually proved much worse.

For LA and many other DCs the key strategic trade-policy issue is how to

invest effectively in product-diversification in order to re-direct export growth from

quadrants 2 to 3. Of course, the design of these policies is sometimes difficult and

their practical implementation is always so, due to institutional, technological,

financial and other constraints. The movement from 2 to 3 relates not only to the

critical supply-side differences between concentrating exports in basic unprocessed

primary commodities rather than in products higher up in the value-chain, but also to

the rather obvious fact that if for a DC exports are to be the engine of growth in an

export-led strategy, it is better to be attached to a locomotive (world demand) with

effective pulling-power (demand-growth). This could prove, for example, the

difference between international trade being a positive- or zero-sum game vis-à-vis

competitors.

Export-led growth in post-reform LA has been, at best, a mixed blessing.

Although ‘maquila’ exports have absorbed a significant amount of labour, they have

been characterized by little or no productivity growth, and associated with little or no

36

productivity growth in the rest of the economy. In turn, in the few cases of rapid

productivity growth in unprocessed commodity exports (e.g., some mining and

agricultural products), there is evidence that this has been more a ‘one-off’ type of

phenomenon than a process that could be sustained over time. Moreover, this

productivity growth has had little ‘pulling’ effect on productivity-growth in the rest of

the economy. And from the point of view of command-GDP-growth, at least until the

cyclical upswing in commodity terms of trade that started with 9/11, when exports

were adjusted by the price of imports, the capacity of export-growth to add to the

purchasing power of GDP proved equally disappointing.

Using Krugman’s terminology, post-reform LA may be a paradigmatic case in

which the single-minded emphasis on improved export competitiveness has led policy

makers and their economic advisers to forget that what really matters from the point of

view of income in export-led economies are the purchasing power gains arising from

specialization. And as Krugman emphasizes, single-minded emphasis on export

competitiveness is a very dangerous obsession that tends to skew domestic economic

policies.41

That is not to say that in open economies ‘competitiveness’ does not matter; but

when the terms of trade are so unstable and downwardly-mobile, “competitive-gains”

can mean something very different from “welfare-gains”. Therefore, being able to

enhance international competitiveness in a given product, even when both the domestic

and the international prices are “right”, does not per se justify the continuous

concentration of resources in that specific export product in the long run. In fact, one

should never forget that when Korea took the decision to ‘re-invent’ its export sector in

the early 1960s, it certainly had no ‘competitiveness’ problem with its world class silk

or its excellent quality seaweed (the country’s main export products at the time).

37

LA’s underinvestment in export-product diversification in a world of rapidly-

changing demand has led the region from being a major player in major products to

being a major player in relatively marginal products.42 In fact, it has not even sought

‘niches’ in segmented commodity markets that have emerged from the dissemination

of the new technological paradigm (characterized by a ‘knowledge-society’ with

flexible production techniques). This underinvestment in part reflects the region’s

(path-dependent) obsession with traditional products that are often less and less

attractive from the point of view of international demand, in which further

‘competitive’ gains can only be achieved by ever increasing mass production, more

and more optimal routines, greater and greater degrees of standardization, and larger

and larger economies of scale.43 However, this underinvestment also reflects the fact

that every time private investment in the region manages to rise above 15% of GDP

the capitalist élite begins to experience feelings of vertigo. As a result, LA ends up

being like a football team that plays better and better in a league that gets continuously

relegated to lower and lower divisions, while EA seems to be a member of a league

continually promoted to higher and higher divisions. In fact, in a Ricardian sense,

LA’s increased export competitiveness has sometimes acted as an ‘own-goal’ because

its standard of living has sometimes actually declined when domestic output growth

has been outweighed by deteriorating terms of trade.

The obsession with increased competitiveness in existing export products has

led LA to keep shooting at a fixed target, even though world demand is a rapidly

moving one. In contrast, EA, aided by its own efforts in institution-building,

investment, savings, education, etc., as well as by Japanese regional leadership, has

been increasingly successful in learning how to hit a moving target. Further, as the

data have shown, international demand-adaptation and supply-upgrading are closely

38

related phenomena; therefore, predicting the trajectory of this moving target does not

seem to be such a difficult task, despite the rapidly changing technological paradigm.

Unfortunately, the capacity of the state to ‘discipline’ the capitalist élite in order to be

able effectively to implement an appropriate set of strategic trade and industrial

policies is another matter – and one in which LA does not seem to have learnt how to

acquire much comparative advantage!

One of the characteristics of a great deal of the trade literature, even in some of

its ‘new-trade’ brand, is that it has been slow to acknowledge this type of problem.

For example, when ‘new-trade’ theorists criticize traditional neo-classical trade theory,

they do so only in relation to issues such as the fact that it has ignored the possibility

that a given set of “right” prices could still generate multiple equilibria due to

uncertainties in the economic system (where decisions are taken in a decentralized

way); or because it ignores the possibility of the emergence of several types of

inefficiency in the social learning process, which in turn could produce inefficient

economic outcomes. These could occur, for example, because the information

structure is determined by the nature and the extent of these uncertainties, the degree

of communication among agents, the pay-off relevance of the decisions to be taken and

the actual range of choices faced by the agents. In other words, in these types of

criticism of the neo-classical trade theory, ‘new trade’ theory has often tried to

identify, and give due relevance to, issues such as the social learning process and the

macro-institutional framework of the export sector. It has put emphasis on the actual

context in which agents are supposed to learn, and on how this learning is very much

determined by the institutional context. Further, it has endogenised the learning

process to a (supposedly exogenous) institutional context, and discussed how in this

way different export paths could arise. However, the concept of an “optimal” export

39

path, even though more relative, is still determined by how to maximize an export

performance given present resource endowments and a set of contemporary (more or

less) “right” prices. But it is not at all obvious how the given set of prices will signal

medium- and long-term changes in the pattern of international demand and

technological change, nor how these prices, or the supposedly ‘exogenous’ domestic

institutional setting, can provide an incentive for diversification of exports towards a

flexible pattern compatible with inter-temporal maximization of gains from

specialization. In a sense, this new literature returns to the same traditional issue of

how to achieve the maximum degree of ‘competitiveness’ given some existing

domestic endowments and current external constraints (now including market failures),

but with a much richer way of understanding the dynamics of how this could (or could

not) be achieved.

An alternative approach to international trade in DCs should promote the idea

that once agents and institutions have learnt how to deal effectively in the ‘real’

international markets, there is still a need for some vigorous agency (e.g., active trade

and industrial policies) steering the economy towards a flexible export productive

capacity, which can maximize returns from (often acquired) comparative advantages

in a world with constantly changing demand patterns and rapid technological change.

In this way, a country could attempt to maximize the trade-induced output-growth

potentials it gets from an export-led growth strategy as well as welfare gains from

international trade. As Stiglitz has repeatedly demonstrated, markets do not necessarily

lead to these sorts of efficient outcomes by themselves.44

There are obviously many other trade-related issues that cannot be discussed

here, such as the crucial one of exhaustible resources (even for resource-rich

countries); the optimal taxation policy (“royalties”) towards the extraction of non-

40

renewal export-resources; FDI’s anti-Latin American bias in the local processing of

resources; the upgrading and “anchoring” of maquila production lines; the perennial

lack of innovative Schumpeterian spirit of the Latin American oligarchy; and, in

particular, the question of why the region’s capitalist élite seem to get altitude sickness

every time private investment climbs to about 15% of GDP.

In sum, the remarkable asymmetries in terms of gains from specialization

between LA and EA (evident, for example, in the negative and significant Latin

American dummy in Figure 8.12 above) inevitably lead to a question asked in the

spirit of Sugar Kowalczyk (Marilyn Monroe’s character in Some Like It Hot): why is it

that in trade-related issues Latin American countries always seem to end up with the

wrong end of the lollipop? There can be little doubt that at least one important part of

the answer is that the Latin American capitalist élite can ‘afford’ to be so risk-averse in

terms of demand-adaptation and supply-upgrading in their trade-related business

because they are as effective now (after economic reforms) as they were before in

creating structures of property rights and incentives from which lavishly large income

streams (rents) can be generated in the domestic economy. After all, according to the

World Bank statistics, the median income share of the top 10% of income earners in

LA is currently above 45% (and increasing), while in Malaysia and Thailand it is 35%,

in India 31%, and in Korea only 25%!45

In other words, in post-reform LA the capitalist élite can ‘afford’ to be so risk-

averse in their trade-related business because they have managed to create a political-

institutional settlement in which a new distributional coalition has succeeded in

imposing a new structure of property rights and incentives that have allowed them, and

their populist allies, extravagant forms of predatory capitalism, un-productive rent-

seeking and the economic emasculation of the state.46 And in terms of their populist

41

allies, one question that begs an answer is why democratic forces in the region were so

prepared to accept such a strait-jacketed type of institutionality in exchange for the

return to democracy – the democratic forces in Chile seemed to believe that they could

only defeat Pinochet in his 1988-plebiscite if they implicitly agreed that their electoral

motto would be “Liberté, Inégalité, Passivité”.47 Another, of course, is why almost

immediately after the return to democracy did these formerly progressive groups

decide to become the new standard-bearers of the neo-liberal orthodoxy?48

So, it seems that in their trade-related business the Latin American capitalist

élite can perfectively well ‘afford’ not to make the required investment efforts for

adaptation and upgrading, not to be Schumpeterian-innovative, or to take the required

risks, because they can count on the fact that their (not very puritanical) share of

domestic income will be so plentiful that it will compensate for their (not very

Calvinistic) attitude towards trade-related issues. In fact, if one compares the above-

mentioned share of national income appropriated by the top 10% with the share of

private investment in GDP, LA appears even more as an outlier: while in LA the

income-share of the top decile is approximately three times the GDP-share of private

investment, in India this ratio is less than 1.5, in Malaysia and Thailand is only slightly

above one, and in Korea these shares of income and investment are roughly the same.

Perhaps this is what really makes the discreet charm of the Latin American bourgeoisie

so unique!

42

Appendix 1

A demand-adaptability index vis-à-vis (a changing) OECD structure of imports.

One set of statistics that can be constructed using the data provided by the Trade-CAN

software is a ‘demand-adaptability index’ of a country to the (changing) structure of

imports of the OECD (see Figure 8.4 and footnote 14 above). This index can give an

insight into a country’s capacity to react to changes in OECD demand for imports.

The ‘demand-adaptability index’ jA of a country ‘j’ to the structure of OECD

imports is defined here as ndd i

ndij

i

dijj aaA . The numerator

di

dija is the country's

market share in demand-dynamic sectors (‘d’), adjusted by the weight of demand-

dynamic sectors in OECD imports; and the denominator ndi

ndija is the country's

market share in non-dynamic sectors (‘nd’), also adjusted by the share of these sectors

in OECD imports. Therefore, j

ijij S

sa ; where

i

ijij M

Ms

is the country ‘j’

share ‘s’ in OECD imports (‘M’) of product ‘i’ (for either ‘d’ or ‘nd’ products), and

S M sj ji

ij is the country’s total share ‘S’ in OECD imports (‘M’) – this, in turn, is

the sum of the market shares for different products ‘i’ of country ‘j’. Finally, demand-

dynamic products (‘di’) are those import products that increase their shares in OECD

imports between two points in time (such that 0 idi ss ). In turn, non-dynamic

products are those that decrease their shares in total imports during this period

( 0 indi ss ).

43

Appendix 2

The four quadrants of Figure 8.10

For the export profile of a country to be located in Quadrant 1 during a given period of

time – i.e., for it to be classified as an ‘uncompetitive’ country exporting ‘non-demand-

dynamic’ products during this period (e.g., LA in the 1960s) – its export profile should

be characterized by both [Mi/M (fy) < Mi/M (by)], and [Mij/Mi (fy) < Mij/Mi (by)]; where

‘fy’ is final year and ‘by’ is base year of this period (note that data for all years in the

Trade-CAN software are 3-year moving averages). For an export profile to be in

‘Quadrant 2’ – i.e., a ‘competitive’ country exporting ‘non-demand-dynamic’ products

(e.g., LA in the 1990s) – this profile should be characterized by [M

t

i/M (fy) < Mi/M (by)]

and [Mij/Mi (fy) > Mij/Mi (by)]. For it to be in ‘Quadrant 3’ – a ‘competitive’ country

exporting ‘demand-dynamic’ products (e.g., EA in the 1990s) – the conditions [Mi/M

(fy) > Mi/M (by)], and [Mij/Mi (fy) > Mij/Mi (by)] should apply. Finally, to be in ‘Quadran

4’ – and be an ‘uncompetitive’ country exporting ‘demand-dynamic’ products (e.g.,

Japan and the European Union in the 1990s) – a country’s export profile should be

characterized by [Mi/M (fY) > Mi/M (by)] and [Mij/Mi (fy) < Mij/Mi (by)].

44

Bibliography

Akamatsu, K (1962), ‘A Historical Pattern of Economic Growth in Developing Countries’; The Developing Economies, 1(2).

Amsden, A (2001), The Rise of “the Rest”. Challenges to the West from late-industrializing economies; OUP.

Bhagwati, J (2004), ‘Testimony to the U.S. House of Representatives Committee on Financial Services Subcommittee on Domestic and International Monetary Policy, Trade and Technology’, April 1

Blankenburg, S, J G Palma and F Tregenna (2008), ‘Structuralism’, The New Palgrave Dictionary of Economics, 2nd Edition, Macmillan.

Capdevielle, M (2005), ‘Procesos de producción global : ¿alternativa para el desarrollo mexicano?’, Comercio Exterior, 55(7).

Chang, H-J (1994), The Political Economy of Industrial Policy, Macmillan.

Chang, H-J (2007), Bad Samaritans: The Myth of Free Trade and the Secret History of Capitalism, Random House.

Cimoli, M (1994), Lock-in and specialization (dis)advantages in a structuralist growth model, in J Fagerberg, N Von Tunzelmann and B Verspagen (eds.) The Dynamics of Technology, Trade and Growth, Elgar.

Cimoli, M and N Correa (2005), ‘Trade openness and technology gaps in Latin America: a “low-growth trap”?’, in J A Ocampo (2005).

Diaz-Alejandro, C (1984), ‘Latin American Debt: I don't think we are in Kansas anymore’, Brookings Papers on Economic Activity, 2.

DiJohn, J (2008), From Windfall to Curse? Oil and industrialisation in Venezuela since 1920, Pennsylvanian State University Press.

Dosi, G, C Freeman, R Nelson and L Soete (1988), Technological change and Economic Theory, Pinter.

Dosi, G, K Pavitt and L Soete (1990), The Economics of Technical Change and International Trade, Harvest.

Ffrench-Davis, R, O Muñoz and J G Palma (1994), ‘The Latin American Economies, 1950-1990’, Cambridge History of Latin America, 6, CUP.

Freeman, C, and L Soete (1997), The Economics of Industrial Innovation, Pinter.

Hobday, M (2003), ‘Innovation in Asian Industrialisation: a Gerschenkronian perspective’, Oxford Development Studies, 31(3).

Kahn, M (2000), ‘Rent-seeking as Process’, in M Khan and KS Jomo (eds.), Rents, Rent-Seeking and Economic Development: Theory and Evidence in Asia, CUP.

Krugman, P (1997), Pop Internationalism; MIT Press.

Marcel, M and J G Palma, (1988), ‘Third World debt and its effects on the British economy: a Southern view on economic mismanagement in the North’, Cambridge Journal of Economics, 12(3).

Mayer, Jorg (2003), ‘The fallacy of composition: A review of the literature’, UNCTAD Discussion Papers 166.

McKay, J (2002), Export Structures in a Regional Context and the Flying Geese Paradigm; Ph D Thesis, University of Cambridge.

Moguillansky, G (1999), La Inversión en Chile: ¿el fin de un ciclo en expansión?; Fondo de Cultura Económica.

Moreno-Brid, J C, J Santamaría, and J C Rivas Valdivia (2005), ‘Industrialization and Economic Growth in Mexico after NAFTA: The Road Traveled’, Development and Change, 13(6).

Ocampo, J A and M A Parra (2004), ‘The commodity terms of trade and their strategic implications for development’; ECLAC.

Ocampo, J A and M A Parra (2006), The dual divergence: growth successes and collapses in the

45

developing world since 1980, DESA.

Ocampo, J A (2004), ‘LA in the long 20th Century’, UN, mimeo.

Ocampo, J A (2005), Beyond Reforms, Structural Dynamics and Macroeconomic Vulnerability, Stanford University Press.

Palma, J G (1998), ‘Does it make a difference to export micro-chips rather than potato-chips? Comparing export structures in East Asia and Latin America', UNCTAD.

Palma, J G (2003), ‘National inequality in the era of globalisation: what do recent data tell us?’, in J Michie (ed.), Handbook of Globalisation, Elgar.

Palma, J G (2004), ‘Latin American during the second half of the 20th Century: from the ‘age of extremes’ to the age of ‘end-of-history’ uniformity’, in H-J Chang, Rethinking Development Economics, Anthem Press.

Palma, J G (2005a), ‘The six main ‘stylised facts’ of the Mexican economy since trade liberalisation and NAFTA’, Journal of Industrial and Corporate Change, 14(6).

Palma, J G (2005b), ‘Four sources of de-industrialisation and a new concept of the Dutch Disease', in J A Ocampo (2005).

Palma, J G (2006), ‘Stratégies actives et stratégies passives d’exportation en Amérique Latine et en Asie orientale. La croissance en tant que phénomène spécifique á la composition de produit et spécifique aux institutions’, Revue Tiers Monde, 186.

Palma, J G (2008a), ‘Raúl Prebisch’, The New Palgrave Dictionary of Economics, 2nd Edition, Macmillan.

Palma, J G (2008b), ‘Why did the Latin American critical tradition in the social sciences become practically extinct?, in M Blyth (ed.), The Handbook of International Political Economy, Routledge.

Pérez, C. (2002), Technological Revolutions and Financial Capital, Elgar.

Pérez, C. (2008), ‘A vision for Latin America: a resource-based strategy for technological dynamism and social inclusion’, ECLAC.

Pesaran H, N U Haque and S Sharma (2000), ‘Neglected heterogeneity and dynamics in cross-country savings regressions’, in J Krishnakumar and E Ronchetti, (eds.) Panel Data Econometrics - Future Direction, Elsevier Science.

Lagos, J. F. (2000), Chile, de Exportador de Cobre Refinado a Gran Exportador de Concentrados: ¿la segunda fase exportadora?, mimeo,

Stiglitz, J (2004), ‘The Free Trade Agreement between Chile and the US’, mimeo.

Stiglitz, J (2002), Globalization and its Discontent; Penguin.

Teece, D (1996), Strategy, Technology and Public Policy: The Selected Papers of David Teece, Vol. 2, Elgar.

Trade-CAN (2005), Competitive Analyses of Countries; ECLAC-World Bank.

Trần Văn Thọ, Nguyễn Ngọc Đức, Nguyễn Văn Chỉnh, Nguyễn Quán (2000), The Vietnamese Economy 1955-2000. New Calculations and New Analyses, Statistical Publishing House - Hanoi.

Tregenna, F (2006), ‘Towards a new understanding of Botswana’s growth experience’, paper presented at the AEA meeting, Boston.

UNCTAD (1996), Trade and Development Report, UN.

Walker, M (2003), ‘Resource-based industrialisation strategies: lessons from the developed world experience’, mimeo.

WDI (2007), World Development Indicators, World Bank.

46

47

1 I would like to thank Alice Amsden, Stephanie Blankenburg, Mario Cimoli, Mushtaq Khan, José Luis Fiori, Daniel Hahn, Michael Hobday, Richard Kozul-Wright, Carlos Lopes, Carlota Perez, Annalisa Primi, Bob Sutcliffe, Fiona Tregenna and especially Giovanni Dosi and Julie McKay for helpful observations. Also, participants at seminars in Bangkok, Bilbao, New Delhi, Kuala Lumpur, Rio de Janeiro, Santiago and Sydney made constructive comments on this paper and on two previous drafts (Palma, 1998 and 2006). The usual caveats apply. 2 If the rates of growth of GDP pc of the last five years in both countries were to continue, China would only need another 18 years to catch up with Mexico. 3 See Ffrench-Davis, Muñoz and Palma (1994) and Palma (2004). For an overview of Latin American economic development throughout the 20th century, see Ocampo (2004). 4 For example, the combination of the Fed trebling nominal rates and recession in the North led to a 33 percentage-point increase in the real rate of interest paid by LA for its foreign debt (i.e., when LIBOR is deflated by LA’s non-oil export-prices, this rate increased from minus 11.2% to 22.1%; see Palma, 2004). See also Diaz-Alejandro (1984), Marcel and Palma (1988) and Ocampo (2004). 5 The new economic discourse went as far as transforming almost anything previously considered as virtue into vice, and vice-versa. This attitude was best summarized by Gustavo Franco, President of the Central Bank during the years of economic reform (leading up to the 1999 financial crisis): “[Our real task] is to undo forty years of stupidity (‘besteira’) […].” (Veja 15-11-1996) 6 Franco again: “[In Brazil today] the alternative is to be neo-liberal or neo-idiotic (neo-burros).” 7 See, for example, McKay (2002) and Palma (1998). 8 With the exception of post-NAFTA Mexico, this openness was clearly not extended to LA, and especially not to Brazil. 9 This also led to remarkably positive interaction between increases in productivity and wages, particularly in the first-tier NICs (Newly Industrializing Countries). 10 The price of memory chips (per MB of RAM), for example, fell by about four-fifths in 1996 alone, partly resulting from excess supply due to massive expansion of Taiwanese production. Korea was particularly affected. 11 This was reflected in rising debt/equity ratios, which, particularly in Korea, reached vertiginous heights. 12 See Ocampo and Parra (2004), and Palma (2008a). 13 See Blankenburg, Palma and Tregenna (2008). 14 This long-term trend has been reversed since 9/11, the Iraq war and the rise of China and India, at least in oil and in some minerals and agricultural products. As has so often happened in the past, many policy-makers and analysts have incorporated a bit too quickly into their analysis the assumption that this post-9/11 cyclical recovery is a permanent state of affairs. 15 See Krugman (1994). 16 Paraphrasing Canon’s ad, EA’s attitude towards exports could be summarized as “if anybody can, we can!” 17 See Appendix 1. 18 In 2000, for example, 54% of OECD imports were in sectors that had been “demand-dynamic” between 1980 and 2000; so an index of 1 for a country in 2000 means that its own market shares in OECD imports in 2000 had the same structure (54%/46%). 19 For a critique of the ‘resource curse’ literature, see DiJohn (2008). For a detailed study of how one particularly successful commodity exporter (Botswana) managed to avoid this supposed ‘curse’, see Tregenna (2006). 20 On the literature on ‘technology-gap’ and evolutionary theory, see Cimoli (1994); Dosi et al. (1988)

48

and (1990); Freeman and Soete (1997); Hobday (2003); and Teece (1996). 21 One example of this would be organic agriculture (see Perez, 2008). 22 Sweden had a similar capacity to supply-upgrade its timber exports, as well as the production of inputs for its commodity-extracting industries (see Walker, 2003). 23 Taiwan presents a similar success story of upgrading in the timber-processing industry, but this country had to import the required inputs. 24 On the technological issues involved, see especially Perez (2003). 25 In fact, the first car that Toyata exported to the US (the ‘Toyopet’) was such a flop that it had to be withdrawn from the market (it was described at the time as just “four wheels and an ashtray”). As Chang reminds us, “many [inside and outside Japan] argued at the time that Toyota should have stuck to its original business of making simple textile machinery. […] Today, Japanese cars are considered as ‘natural’ as Scottish salmon or French wine.” (2007). 26 On the distinction between “allocative”, “Schumpeterian” and “growth” efficiencies, see Dosi, Pavitt and Soete (1990) and the chapter by Cimoli, Dosi, Nelson and Stiglitz in this book. 27 See especially Amsden (2001). 28 Sometimes Japan ends up only relinquishing its productive capacity in some inputs of a given product, a process that leads to a regional integration of production lines. 29 In this context, ‘the most dynamic sectors’ means the sectors in which the US had the highest increase in OECD market shares. 30 Not surprisingly, in a series of new trade agreements between the US and Latin American countries, a primary concern of the US has been the opening up of the economies of Latin American countries to its exports of primary commodity (particularly agricultural products). 31 The main limitation of traditional trade statistics (such as those available in the UN Com-trade online database and the Trade-CAN software) is that they refer to gross value of exports and not to their value added. As mentioned above, this problem is particularly important in the study of the so-called ‘maquila’ industry (or purely assembly-type export-operations) of Mexico and some parts of Central America; in fact, the use of these data would distort the comparative regional analysis attempted below. As a result, there is little option but again to exclude Mexico in most of the following comparative analysis, and to analyze Mexico’s peculiar export-led growth experience properly separately (see Palma 2005a). 32 The partial exception is Brazil, which increased its share of exports in ‘high-tech’ products from 0.3% in 1963 to 3.8% in 1985 and 9.4% in 2000. However, two thirds of those exports in 2000 consisted only of only one (though rather notable) product: aircrafts (SITC group 792). In fact, in that year no other ‘high-tech’ product reached 1% of total exports. 33 For an analysis of the fallacy of composition issue, see Mayer (2003). 34 See especially Kahn (2000). 35 On the issue of the different “qualities of specializations”, see Cimoli and Correa (2005). 36 On this subject, see especially Lagos (2000). On the issue of Chile’s and other Latin American countries’ “premature” de-industrialization since economic reform, see Palma (2005b). 37 See especially Moguillansky (1999). 38 Further, the proliferation of the above-mentioned trade agreements between countries of the region and the US (such as those signed by Chile in 2003, and by some Central American countries in 2004), especially due to their asymmetric nature (see, for example, Stiglitz, 2003, and Bhagwati, 2003), would probably make the transition from quadrant 2 to 3 even less likely, unless one considers (against all evidence – see Palma, 2005a) that maquila-type exports are the panacea. 39 One of the most remarkable features of Mexico’s poor growth performance is that it took place in a context of both a massive inflow of FDI (well over US$ 200 billion during this period; see Palma, 2005a) and practically unrestricted market access to the US – the first two items on most DCs’ growth

49

agenda! 40 This extremely slow rate of per capita GDP growth contrasts with that achieved by Mexico during its previous politically “populist”, financially “repressed”, and economically “distorted” four decades (1940-81). During those years, Mexico (with a much faster rate of population growth) achieved a rate of GDP per capita growth of 3.5% per annum. There is little doubt that at the time of its 1982 financial crisis, Mexico needed major political, economic and institutional re-engineering; however, the chosen path does not seem to have been the most effective. In fact, if before 1980 Mexico doubled its income per capita every 20 years, at the current pace it would take more than 70 years to do the same! 41 See especially 1997, chapters 1 and 2. 42 In 1963, 57% of LA’s exports were products included in the OECD list of 20 major import products of the time; by 2000, less than 7% of its exports consisted of products that made the then (new) OECD ‘top 20’ list. Meanwhile the first- and second-tier NICs and China moved in the opposite direction, with respective shares of 6% and 64%, 8% and 50%, and 8% and 32%, respectively. 43 In 2000, for example, for all of Chile’s remarkable export growth and (horizontal) diversification, two thirds of its exports still consisted of products that belonged to the 1963 list of ‘top-20’ OECD imports, while only 2.6% of its exports appeared in the 2000 ‘top-20’ list. 44 See, for example (2002), and the introduction to this book by Dosi, Cimoli and Stiglitz. 45 For a comparative regional analysis of income distribution, see Palma (2003). 46 For example, the processes of privatization of public assets in the region compete in their degrees of corruption with those of post-communist Russia. At the same time, the post-reform institutional environment has created a situation in which large corporations in the real and financial sectors can exploit their monopoly power freely, and can count on unlimited bail-outs from governments when things go wrong. In this institutional environment capital can also rent-seek freely from downwardly-flexible labour markets, and can be assured that their capital flight will not only be free of regulation but will also be highly exchange-rate-subsidized (for example, foreign asset holdings by Argentinean residents at the time of the 2001 crisis amounted to about $ 100 billion; this figure was 25% higher than the whole increase in foreign debt since the first election of Menem in 1989). 47 The Latin American economic élite’s capacity to transform particularly unequal forms of income distribution achieved during repressive regimes into permanent features of society has become legendary. 48 For an analysis of these issues, see Palma (2008b).